HIREX Pointing Control and Image Stabilization
Overview:
The stated HIREX imaging goal is resolution at the 0.01 arcsec level. There
are many components that affect the achievement of this goal, one of the
most basic is the stability of the image on the focal plane during an observation.
Responsibility for achieving this goal is distributed between the instrument
and the spacecraft. The present plan is to permit the spacecraft up to
5 arcsec of pointing error. Image motion within that band is removed by
a tip/tilt servo system on the secondary. The primary image motion sensor
is a set of correlating image trackers mounted at the focal plane around
the science CCD.
Models:
In order to determine the likely image stability on orbit a dynamics simulation
of a full spacecraft model has been made. In order to construct this
model, several intermediate models and modeling systems have been
used. The dynamics model was constructed in AUTOSIM, a dynamics modeling
software package written at the University of Michigan Transportation Department.
This package takes a description of the geometric and constrain layout
of the system and produces a set of equations describing the system dynamics.
The output can be in a number of forms, we have selected an output in MATLAB
form. We have used SDRC IDEAS, both solids modeling and FEA to determine
the system inertial and elastic properties. Finally the dynamics model
was placed within a Simulink model (a block diagram modeling system operating
under MATLAB) where all the control system design and simulation work was
done.
The model includes:
-
The first 2 boom modes,
-
The TXI, and its movement freedom,
-
The angular adjustablity of the secondary,
-
The filter wheel, and its movement freedom,
-
3 mutually orthogonal reaction wheels and related noise,
-
The image stability system and associated control.
Where possible real world effects like the averaging nature of the correlating
star tracker and its sample and hold operation have been included. In later
runs we will examine the effects of limited resolution in both the image
measurement and in the secondary position control.
Preliminary Results:
The preliminary results are promising. The graphs below show aspects of
the predicted system behavior in response to a candidate observatory repointing.
The spacecraft is commanded to change its line of sight by 120 arcsec over
a period of 80 seconds. The pointing path and rate are specified. The resulting
image motion is shown in figure 1.
Figure 1 Image motion During 120 arcsec move
Figure 1 shows that the image remains within the stability requirements
even during this large system repointing. The spike at 80 seconds results
from the details of the pointing command at the end of the move. Specifically
the pointing rate command is set to zero just as the spacecraft overshoots
its intended target, this results in a large reaction torque. The deflection
is the result of fact that the secondary is out on the end of the swinging
boom and thus has an applied inertia torque. This can be overcome with
a more complex control law.
The spacecraft pointing performance is shown in figure 2:
Figure 2 Spacecraft Pointing Error During 120 arcsec Maneuver
The spacecraft, at least the instrument bay, follows the pointing command
quite well. However as Figure 3 shows the boom has a much larger amplitude.
Figure 3 Boom Deflection in Response to a 120 arcsec Maneuver
Please note that this graph shows the deflection of the boom, the scale
on the left hand side of the graph, and the linear decenter of the secondary
at the right. Though boom angle and secondary decenter are not strictly
related by a constant ratio, it is close enough for preliminary performance
determination.
Finally figure 4 shows the reaction torque required to make the simulated
move:
Figure 4 The Commanded Torque During the 120 arcsec Maneuver
It is clear, that at least for the simulated move the reaction torque requirements
are modest and lie within the capability of available reaction wheels.
Questions, comments pcheimets@cfa.harvard.edu